Intelligent robot for bulkhead contamination identification and cleaning and bulkhead cleaning method

Abstract

The application discloses an intelligent robot for bulkhead contamination identification and cleaning and a bulkhead cleaning method. The robot body is equipped with a positioning navigation system, a track assembly, a magnetic adsorption wheel, a mechanical arm assembly and a cleaning brush assembly. The positioning navigation system combines inertial navigation, visual SLAM and graph neural network loop detection to correct errors. The visual unit uses deep learning technology to identify contaminants. The cleaning method is as follows: positioning mapping and image capturing, extending the camera as needed, moving and fixing the robot, adapting the cleaning speed, real-time position correction, and deciding whether to continue after secondary detection. The robot can adaptively adjust the posture and efficiently complete the bulkhead cleaning.

Description

Technical Field

[0001] This invention relates to the field of underwater robots, and in particular to an intelligent robot and a method for cleaning contaminants on the bulkhead. Background Technology

[0002] With the rapid development of marine engineering, ship maintenance, and underwater facility management, the long-term underwater operation of cabin walls and enclosed structures makes them susceptible to the adhesion of marine organisms, corrosion products, and other pollutants. These pollutants not only affect the corrosion resistance of the cabin structure but may also affect the normal operation of sensors and equipment, thereby reducing the overall operational efficiency and safety of the platform. Therefore, conducting efficient and low-loss underwater bulkhead cleaning operations is of significant engineering importance.

[0003] Traditional underwater cleaning operations rely heavily on divers or remotely controlled underwater robots for mechanical brushing, spraying, or chemical treatment. However, the confined space, complex structure, and significant water flow interference in these areas limit manual operation and the mobility of traditional ROVs, resulting in low efficiency and high risks. In such enclosed underwater environments, the inability to receive GPS signals makes robot positioning highly dependent on visual SLAM and inertial navigation systems (INS). However, the former is prone to failure in low-light, blurry, and feature-lacking underwater environments, while the latter suffers from cumulative drift, making it difficult to achieve continuous and stable precise positioning, thus resulting in insufficient autonomous operation capabilities. Furthermore, the contaminants on the confined spaces are diverse and unevenly distributed. Traditional methods often employ indiscriminate cleaning, failing to adaptively match contaminant identification with cleaning intensity, easily leading to over-cleaning, resource waste, and even damage to the confined space coating. Therefore, there is an urgent need for an underwater confined space cleaning robot system that integrates high-precision underwater positioning, intelligent contaminant identification, and adaptive cleaning capabilities to meet the practical needs of automated, efficient, and safe cleaning of confined underwater confined spaces. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent robot for identifying and cleaning bulkhead contaminants, used for automated contamination identification and efficient cleaning of ship bulkheads in underwater environments, solving the problems of low efficiency, difficult positioning, and incomplete cleaning associated with traditional manual cleaning. The specific solution is as follows:

[0005] An intelligent robot for identifying and cleaning bulkhead contaminants includes a robot body equipped with a positioning and navigation system, and the robot body is equipped with a track assembly, magnetic adsorption wheels, a robotic arm assembly, and a cleaning brush assembly.

[0006] The track assembly includes drive tracks mounted on both sides of the bottom of the robot body. The drive tracks include a drive wheel, a driven wheel, and an annular track that is connected between the drive wheel and the driven wheel. The upper and lower surfaces of the annular track protrude from the upper and lower surfaces of the robot body.

[0007] The magnetic adsorption wheel is installed on the tail side of the robot body between the two transmission tracks. The rotation direction of the magnetic adsorption wheel is the same as that of the drive wheel. The magnetic adsorption wheel includes a magnetic ring and a magnetic ring driver for driving the magnetic ring to rotate.

[0008] The robotic arm assembly includes support arms symmetrically arranged on the left and right sides of the robot body. Each support arm includes a robotic arm base, a support arm, and a support arm 1. The robotic arm base is fixed to the robot body. One end of the support arm is movably connected to the robotic arm base, and the other end is rotatably connected to the support arm 1. The end of the support arm 1 is equipped with an electromagnetic chuck. The support arm drives the support arm 1 to rotate up and down, and the electromagnetic chuck provides auxiliary support for the robot body.

[0009] The cleaning brush assembly includes a cleaning arm, a brush head connecting rod, and a disc brush. The cleaning arm is fixedly connected to the brush head connecting rod, and the disc brush is mounted on the brush head connecting rod. The disc brush is equipped with a brush head motor and a disc brush head that is detachably mounted on the rotating shaft of the brush head motor.

[0010] Furthermore, the robot body is equipped with an outer shell, the upper and lower surfaces of which are lower than the upper and lower surfaces of the annular track, and a lifting channel is provided at the top of the outer shell;

[0011] The positioning and navigation system is equipped with a lifting camera, which includes a camera push rod, a connecting rod, a transparent waterproof shell, and a camera. The two ends of the connecting rod are rotatably connected to the camera push rod and the transparent waterproof shell. The transparent waterproof shell slides up and down in the lifting channel. A gimbal motor is installed inside the transparent waterproof shell, and the camera is mounted on the gimbal motor.

[0012] Furthermore, the magnetic ring driver is connected to the cylindrical rotating body, and a magnetic ring with a permanent magnet is fixedly installed on the outer diameter of the cylindrical rotating body.

[0013] Furthermore, the disc brush head is provided with a sleeve that is inserted and connected to the rotating shaft of the brush head motor. The side wall of the rotating shaft of the brush head motor is provided with a radial threaded hole, and the sleeve is provided with a through hole that coincides with the threaded hole.

[0014] Fastening bolts pass through the through holes and connect to the threaded holes to secure the disc brush head to the rotating shaft of the brush head motor.

[0015] Furthermore, the positioning and navigation system also includes a visual positioning unit, which employs a deep learning convolutional neural network and Transformer architecture, combined with multispectral imaging and super-resolution image reconstruction technology, to detect, classify, and assess the degree of contamination of contaminants on the bulkhead.

[0016] Furthermore, the positioning and navigation system combines an inertial navigation system with graph neural network-assisted loop closure detection and map matching functions. It obtains the initial position and attitude estimation through INS, and then uses GNN to perform loop closure detection on visual SLAM map features to correct positioning errors.

[0017] A method for cleaning bulkheads using an intelligent robot, the method comprising the following steps:

[0018] Step 1: The positioning and navigation system is activated to create a map of the current work area. Initial positioning is completed through the inertial navigation system, and the lifting camera is activated to capture images of the bulkhead surface in real time.

[0019] Step 2: The visual recognition module processes the image. If the pollutant coverage is ≥80% or it cannot be determined whether it is a pollutant, the camera push rod pushes the transparent waterproof shell along with the camera to extend along the lifting channel; if the pollutant coverage is <80%, the camera remains retracted.

[0020] Step 3: The track assembly and magnetic adsorption wheels work together to move the robot to one of the pollutant areas. The two support arms extend through the support boom, and the electromagnetic chucks at the ends adsorb onto the cabin wall to achieve auxiliary fixation.

[0021] Step 4: The cleaning boom extends the brush head linkage, and the disc brush head is driven by the brush head motor to rotate and contact the bulkhead. The rotating disc brush head cleans the bulkhead.

[0022] Step 5: During the cleaning process, the positioning and navigation system corrects the position in real time through visual SLAM map and GNN loop closure detection to ensure that no cleaning area is missed;

[0023] Step 6: After cleaning, the lifting camera performs a second inspection. If the contaminants have been removed, the support arm retracts and the robot moves to the next contaminant area; if not, repeat steps 3-5.

[0024] Furthermore, the workflow also includes an attitude adaptive adjustment step: when the robot flips due to the impact of water flow, the magnetic adsorption wheel drives the robot to adaptively flip through the adsorption force of the permanent magnet, so that the track assembly or magnetic adsorption wheel re-adheres to the cabin wall; at the same time, the lifting camera continues to record, and the positioning and navigation system recalibrates its position to ensure that the cleaning work continues.

[0025] Furthermore, the logic of "matching pollutant type with rotation speed" in step 4 is as follows: when the visual recognition module identifies the pollutant as rust through the CNN and Transformer architecture, the rotation speed of the disc brush head is adjusted to 300-500 r / min; when it identifies the pollutant as marine organism attachment, the rotation speed is adjusted to 100-300 r / min.

[0026] The beneficial effects of this invention are as follows: the robot's disc brush head can be rotated by a brush head linkage controlled by a motor; the robot's two robotic arms can be rotated up and down by control rods for auxiliary support and fixation; and the liftable probes on the upper and lower surfaces ensure that the robot maintains accurate recognition and imaging in different environments. The robot's magnetic adsorption wheel assembly uses cylindrical wheels embedded with permanent magnets, providing a strong adsorption effect. Even if the robot flips over, it can continue to work and can adapt to cabin walls with different radii of curvature. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is an external view of an intelligent robot for identifying and cleaning contaminants on the bulkhead according to the present invention.

[0029] Figure 2 This is a structural diagram of a pop-up camera;

[0030] Figure 3 A diagram illustrating the intelligent cleaning robot for the cabin walls, showing it adhering closely to the cabin walls and at an angle to them.

[0031] Figure 4 This is a flowchart illustrating the workflow of the underwater cleaning robot for the outer wall of a ship according to the present invention.

[0032] Figure 5 A flowchart illustrating the working logic of a positioning and navigation system for intelligent robots.

[0033] Figure 6 Flowchart of the working logic for anti-fouling and attitude stabilization control of intelligent robots Detailed Implementation

[0034] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0035] To fully understand this invention, detailed steps and structures will be presented in the following description to illustrate the technical solution of this invention. Preferred embodiments of the invention are described in detail below; however, in addition to these detailed descriptions, the invention may have other embodiments.

[0036] An intelligent robot for identifying and cleaning contaminants on bulkheads includes a robot body 3 equipped with a positioning and navigation system. The robot body 3 has a shell 12, the upper and lower surfaces of which are lower than the upper and lower surfaces of the annular track 10. A lifting channel is provided at the top of the shell 12. The positioning and navigation system includes a lifting camera, which comprises a camera push rod 15, a connecting rod 14, a transparent waterproof shell 2, and a camera 9. The two ends of the connecting rod 14 are rotatably connected to the camera push rod 15 and the transparent waterproof shell 2. The transparent waterproof shell 2 slides up and down within the lifting channel. A gimbal motor is installed inside the transparent waterproof shell 2, and the camera 9 is mounted on the gimbal motor.

[0037] The positioning and navigation system includes a visual positioning unit, which employs a deep learning convolutional neural network and Transformer architecture, combined with multispectral imaging and super-resolution image reconstruction techniques, for detecting, classifying, and assessing the degree of contamination of bulkhead contaminants. The positioning and navigation system integrates an inertial navigation system (INS) with graph neural networks (GNNs) to assist in loop closure detection and map matching. It obtains initial position and attitude estimates through the INS, and then uses a GNN to perform loop closure detection on visual SLAM map features to correct positioning errors.

[0038] In addition, the robot body 3 of the present invention is equipped with a track assembly, a magnetic adsorption wheel 13, a robotic arm assembly, and a cleaning brush assembly.

[0039] The track assembly includes drive tracks mounted on both sides of the bottom of the robot body 3. Each drive track comprises a drive wheel 6, a driven wheel, and an annular track 10 connecting the drive wheel 6 and the driven wheel. The upper and lower surfaces of the annular track 10 protrude from the upper and lower surfaces of the robot body 3. Regardless of whether the robot is on a horizontal, vertical, or inclined bulkhead, the track prioritizes contact with the bulkhead, preventing interference from the main body shell. Simultaneously, the flexible structure of the track can adapt to the curved surfaces or minor protrusions of the bulkhead (such as slight deformation after welding). Combined with the differential speed movement of the two tracks (one track rotates faster than the other), the robot can achieve flexible steering, adjusting its direction even in narrow corners of a compartment.

[0040] The magnetic adsorption wheel 13 is located on the tail side of the robot body 3 between the two drive tracks, forming a "triangular support" structure with the tracks to improve overall adsorption stability. The rotation direction of the magnetic adsorption wheel is the same as that of the drive wheel 6, ensuring that its movement direction is synchronized with the tracks and avoiding motion interference.

[0041] The adsorption wheel is equipped with a magnetic ring and a magnetic ring driver: the magnetic ring driver is a high-precision stepper motor that can precisely control the rotation speed of the adsorption wheel; the magnetic ring is made of permanent magnet material (such as high-performance neodymium iron boron magnets) and is fixed on the outer diameter of the cylindrical rotating body. The upper and lower surfaces of the magnetic ring are flush with the upper and lower surfaces of the annular track 10, which means that when the robot moves, the magnetic ring and the track can simultaneously contact the bulkhead. The track provides the power for movement, and the magnetic ring is firmly attached to the steel bulkhead by magnetic force, preventing the robot from slipping even under the impact of water flow.

[0042] The robotic arm assembly includes support arms arranged on the left and right sides of the robot body 3. Each support arm consists of a robotic arm base 11, a support arm 7, and a support arm 8. The robotic arm base 11 is fixed to the main body with high-strength bolts to ensure overall rigidity. The support arm 7 is connected to the base by a hinge and can rotate up and down under the drive of a motor. The end of the support arm 7 is rotatably connected to the support arm 8, and the support arm 8 can be further adjusted in extension angle, and finally contacts the bulkhead through an electromagnetic chuck at its end.

[0043] When the robot moves to the contaminated area for cleaning, the electromagnetic chuck is energized to generate a magnetic force (the magnetic force disappears after power is cut off, making it easy to retract), firmly adhering to the bulkhead. This auxiliary support serves two purposes: firstly, it counteracts the recoil force generated by the high-speed rotation of the cleaning brush, preventing the robot from swaying; secondly, on inclined bulkheads, it supplements the magnetic attraction force of the magnetic wheels, preventing the robot from tipping over due to a shift in its center of gravity. The rotation angle of the support arm 7 can be adaptively adjusted according to the inclination of the bulkhead (e.g., on a vertical bulkhead, the support arm can extend to a 90° angle with the main body, providing vertical tension).

[0044] The cleaning brush assembly consists of a cleaning arm 4, a brush head connecting rod 5, and a disc brush 1. The cleaning arm 4 is a multi-section telescopic structure driven by a hydraulic or electric push rod, which can adjust the extension distance of the brush head; the brush head connecting rod 5 is rigidly connected to the cleaning arm 4, and the disc brush 1 installed at its end is the component that directly acts on the contaminants.

[0045] The disc brush's structural design features a waterproof servo motor that can output different speeds (adapting to various contaminants). The disc brush head is detachable, with a sleeve that connects to the motor's rotating shaft. The motor shaft has radial threaded holes on its sidewall, and the sleeve has corresponding through holes. A fastening bolt passes through the through hole and connects to the threaded hole to secure the brush head. When the brush head wears (e.g., bristles become shorter) or the brush head type needs to be changed (e.g., a stiff-bristled brush for stubborn rust, a soft-bristled brush for marine life), it can be quickly disassembled and replaced without complicated tools. Furthermore, the rotating disc brush head creates a circular cleaning surface, which, combined with the movement of the cleaning arm, can cover a large area.

[0046] like Figure 1 , 4As shown in Figure 5, the robot's working process is a "production line operation" with the cooperation of various systems. The specific process is as follows:

[0047] After the control system is started, the inertial navigation system (INS) is first triggered to complete the initial positioning. At the same time, the lifting camera is activated. The camera adjusts its angle through the gimbal motor to capture images of the bulkhead in real time and transmits the data to the visual recognition module. The visual recognition module uses CNN and Transformer models to analyze the images: extracting features such as the outline and color of pollutants, and constructing a pollutant distribution map of the local area (e.g., marking area A as having corrosion).

[0048] After localization, visual SLAM continuously updates the environmental map, and GNN performs loop closure detection in real time to ensure the robot always knows "where it is." When a contaminant area is detected, the tracks and magnetic adsorption wheels move in tandem (the tracks provide the main power, and the adsorption wheels prevent slippage), precisely moving the robot to the target area. At this time, the support arm 7 on both sides drives the support arm 8 to extend, and the electromagnetic chuck at the end is energized to adsorb onto the cabin wall, firmly fixing the robot in place.

[0049] Cleaning Phase: The cleaning boom 4 extends until the brush head contacts the contaminant surface (contact pressure can be adjusted via sensors to avoid damaging the bulkhead coating). The brush head motor starts, and the disc brush head rotates at a speed matched to the type of contaminant, removing the contaminant through bristle friction. During the cleaning process, a camera captures real-time footage of the cleaning area. If any area is found to be uncleaned, the cleaning boom will adjust its position to perform additional cleaning.

[0050] After cleaning, the camera performs a second inspection: if the residual area of ​​the contaminant is less than 5%, it is determined to be "removed", the electromagnetic chuck of the support arm is de-energized and retracted, and the robot moves to the next area; if the residual area is ≥5%, the fixing-cleaning process is repeated until the standard is met.

[0051] The camera extends as needed based on the level of contamination: the camera is in a retracted state by default (stored in the lifting channel of the housing), at which time the housing can protect the camera from direct water impact and reduce water resistance when the robot moves.

[0052] After the visual recognition module analyzes the image, it triggers two extension scenarios: First, when the contaminant coverage is ≥80% (e.g., a 20cm×20cm area is almost entirely covered with rust), a clearer and wider field of view is needed to plan the cleaning path (e.g., cleaning by area). The camera push rod 15 pushes the transparent waterproof shell 2 along the lifting channel through the connecting rod 14 to increase the shooting height and range. Second, when it is impossible to determine whether it is a contaminant (e.g., an area with abnormal color but blurry texture), the extended camera can take close-up shots to obtain clearer details (e.g., judging whether it is a stain or a coating defect of the bulkhead itself through texture) to avoid misjudgment.

[0053] When the contaminant coverage is less than 80% and clearly identifiable, the camera remains retracted: firstly, the local cleaning path is clear at this point, eliminating the need to expand the shooting range; secondly, the retracted state reduces energy consumption (reducing energy consumption from push rod actuation and camera resistance exposed to water flow). Furthermore, the robot's outer shell and the camera's waterproof casing are coated with a self-cleaning, anti-fouling coating (such as a fluorinated coating), which reduces contaminant adhesion to the surface, ensuring a clear camera field of view even during long-term operation and reducing the frequency of manual cleaning and maintenance.

[0054] After a work cycle (completing the cleaning of a contaminated area), the robot's movements are seamless, ensuring efficient operation.

[0055] The brush head motor first stops rotating, and the cleaning arm 4 drives the brush head connecting rod 5 to retract (the brush head moves away from the bulkhead to avoid scratching). Then, the visual recognition module performs a secondary inspection of the cleaning area through the camera. If the image analysis shows that the pollutant residue is <5% (considered "removed"), the electromagnetic chucks on both sides of the support arm are de-energized (losing their magnetic force), and the support arm 7 drives the support arm 8 to retract to the initial position (close to the robot body to reduce motion resistance).

[0056] Next, the robot enters the "movement planning" stage: the positioning and navigation system plans the optimal movement path based on the SLAM map and the location of the unwashed contaminants (such as moving in a straight line to the nearest contaminant area to avoid detours); the tracks and magnetic adsorption wheels start working together to move along the planned path.

[0057] If a second inspection reveals that contaminants have not been removed (residual ≥5%), the robot will not move, the support arm will remain fixed, the cleaning arm will extend again, and the brush head will re-clean with adjusted parameters (such as increasing the rotation speed for stubborn contaminants) until the inspection meets the standards. The entire process requires no manual operation, achieving closed-loop automation of "identification-cleaning-re-inspection-re-cleaning".

[0058] When the robot flips over due to wave impact, the magnetic adsorption wheels 13 automatically restore its posture, allowing it to continue working effectively. The robotic arm's active extension and retraction provide additional support, effectively resisting the effects of lateral water flow and preventing tipping. The program logic is as follows: Figure 6 As shown.

[0059] The preferred embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and the devices and structures not described in detail should be understood as being implemented in a conventional manner in the art. Any person skilled in the art can make many possible variations and modifications to the technical solutions of the present invention using the methods and techniques disclosed above, or modify them into equivalent embodiments with equivalent changes, without departing from the scope of the present invention. This does not affect the essential content of the present invention. Therefore, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the present invention's technical solutions still fall within the protection scope of the present invention.

Claims

1. A method for cleaning bulkheads using an intelligent robot for identifying and cleaning bulkhead contaminants, the intelligent robot comprising a robot body (3) equipped with a positioning and navigation system, characterized in that, The robot body (3) is equipped with a track assembly, magnetic adsorption wheels (13), a robotic arm assembly, and a cleaning brush assembly; The track assembly includes a transmission track installed on both sides of the bottom of the robot body (3). The transmission track includes a drive wheel (6), a driven wheel, and an annular track (10) that is connected between the drive wheel (6) and the driven wheel. The upper and lower surfaces of the annular track (10) protrude from the upper and lower surfaces of the robot body (3). The magnetic adsorption wheel (13) is installed on the tail side of the robot body (3) between the two transmission tracks. The rotation direction of the magnetic adsorption wheel (13) is consistent with the rotation direction of the drive wheel (6). The magnetic adsorption wheel (13) includes a magnetic ring and a magnetic ring driver for driving the magnetic ring to rotate. The upper and lower surfaces of the magnetic ring are flush with the upper and lower surfaces of the annular track (10). The robotic arm assembly includes support arms symmetrically arranged on the left and right sides of the robot body (3). The support arm includes a robotic arm base (11), a support arm (7), and a support arm (8). The robotic arm base (11) is fixed on the robot body (3). One end of the support arm (7) is movably connected to the robotic arm base (11), and the other end is rotatably connected to the support arm (8). The end of the support arm (8) is provided with an electromagnetic chuck. The support arm (7) drives the support arm (8) to rotate up and down, and the electromagnetic chuck provides auxiliary support for the robot body (3). The cleaning brush assembly includes a cleaning boom (4), a brush head connecting rod (5), and a disc brush (1). The cleaning boom (4) is fixedly connected to the brush head connecting rod (5), and the disc brush (1) is mounted on the brush head connecting rod (5). The disc brush (1) is equipped with a brush head motor and a disc brush head that can be detachably mounted on the rotating shaft of the brush head motor. The positioning and navigation system also includes a visual positioning unit. The visual positioning unit adopts a deep learning convolutional neural network and a Transformer architecture, combined with multispectral imaging and super-resolution image reconstruction technology, for detecting, classifying, and assessing the degree of contamination of bulkhead pollutants. The positioning and navigation system combines an inertial navigation system with a graph neural network-assisted loop closure detection and map matching function. It obtains the initial position and attitude estimation through INS, and then uses GNN to perform loop closure detection on visual SLAM map features to correct positioning errors. The method includes the following steps: Step 1: The positioning and navigation system is activated to create a map of the current work area. Initial positioning is completed through the inertial navigation system, and the lifting camera is activated to capture images of the bulkhead surface in real time. Step 2: The visual recognition module processes the image. If the pollutant coverage is ≥80% or it cannot be determined whether it is a pollutant, the camera push rod pushes the transparent waterproof shell along with the camera to extend along the lifting channel; if the pollutant coverage is <80%, the camera remains retracted. Step 3: The track assembly and magnetic adsorption wheel (13) work together to move the robot to one of the pollutant areas. The two side support arms extend through the support arm drive, and the end electromagnetic chuck adsorbs the cabin wall to achieve auxiliary fixation. Step 4: The cleaning boom extends the brush head linkage, and the disc brush head is driven by the brush head motor to rotate and contact the bulkhead. The rotating disc brush head cleans the bulkhead. The disc brush head rotates at a speed that matches the type of contaminant. Step 5: During the cleaning process, the positioning and navigation system corrects the position in real time through visual SLAM map and GNN loop closure detection to ensure that no cleaning area is missed; Step 6: After cleaning, the lifting camera performs a second inspection. If the contaminants have been removed, the support arm retracts and the robot moves to the next contaminant area; if not, repeat steps 3-5. The method also includes an attitude adaptive adjustment step: when the robot is overturned by the impact of water flow, the magnetic adsorption wheel (13) drives the robot to adaptively flip through the adsorption force of permanent magnets, so that the track assembly or the magnetic adsorption wheel (13) re-attaches to the cabin wall; at the same time, the lifting camera continues to shoot, and the positioning and navigation system recalibrates the position to ensure that the cleaning work is carried out continuously.

2. The method for cleaning bulkheads using an intelligent robot for identifying and cleaning bulkhead contaminants, as described in claim 1, is characterized in that... The robot body (3) is provided with a shell (12). The upper and lower surfaces of the shell (12) are lower than the upper and lower surfaces of the annular track (10). A lifting channel is provided on the top of the shell (12). The positioning and navigation system is equipped with a lifting camera. The lifting camera includes a camera push rod (15), a connecting rod (14), a transparent waterproof shell (2) and a camera (9). The two ends of the connecting rod (14) are rotatably connected to the camera push rod (15) and the transparent waterproof shell (2). The transparent waterproof shell (2) slides up and down in the lifting channel. A gimbal motor is installed inside the transparent waterproof shell (2), and a camera (9) is installed on the gimbal motor.

3. A method for cleaning bulkheads using an intelligent robot for identifying and cleaning bulkhead contaminants, as described in claim 1, characterized in that... The magnetic ring actuator is connected to the cylindrical rotating body, and a magnetic ring with a permanent magnet is fixedly installed on the outer diameter of the cylindrical rotating body.

4. A method for cleaning bulkheads using an intelligent robot for identifying and cleaning bulkhead contaminants, as described in claim 1, characterized in that... The disc brush head is equipped with a sleeve that is inserted and connected to the rotating shaft of the brush head motor. The side wall of the rotating shaft of the brush head motor is provided with a radial threaded hole, and the sleeve is provided with a through hole that coincides with the threaded hole. Fastening bolts pass through the through holes and connect to the threaded holes to secure the disc brush head to the rotating shaft of the brush head motor.

5. A method for cleaning bulkheads using an intelligent robot for identifying and cleaning bulkhead contaminants, as described in claim 1, characterized in that... In step 4, when the visual recognition module identifies the pollutant as rust using the CNN and Transformer architecture, the rotation speed of the disc brush head is adjusted to 300-500 r / min; when it identifies the pollutant as marine organism attachment, the rotation speed is adjusted to 100-300 r / min.

Images